Abstract:Based on the homogenized surface relative humidity data of more than 2400 national meteorological station of China, and by using TPS (thin plate spline), the 1951-2014 regional ground homogenized relative humidity site data were spatially interpolated, obtaining the Chinese ground homogenized 0.5°×0.5° relative humidity dataset. The quality evaluation results of the dataset show that the interpolation error in winter is relatively higher than that in summer. There are no significant spatial regional characteristics in the interpolation errors of the analysis and observation values in winter, spring, summer and autumn. The corresponding average deviation is 0.002%, 0.013%, 0.008% and 0.007% per month, and the average relative error is 0.431%, 0.439%, 0.286% and 0.382% per month, respectively. The correlation coefficient between the analysis value and the observation value is 0.89. After the interpolation, the gridded relative humidity data can describe the major spatial features of “southeast wet and northwest dry” for the annual average relative humidities of China more accurately and intensively. In addition, it can also reveal the varying features of relative humidity in the regions of the south to Yanytze River the south to Yellow River and north to the Yangtze River, the Northwest China, the northern and southern Tianshan Mountains, Tarim Basin terrain, etc. As there are not enough stations on the Qinghai Tibet Plateau, it is difficult to give a qualitative and quantitative analysis of the characteristics of the relative humidity of the air in this region. The long time series of climate change trend analysis, based on the CR dataset shows that the national average relative humidity in the late 60 years is in a decreasing trend.